Merge pull request #285 from Parallel-in-Time/v6

Proposition for release 5.2

Author: pancetta

CONTRIBUTING.md

@@ -7,12 +7,13 @@ Additionally, a _few_ rules are set to enforce code readability, consistency and
 Others are specific conventions chosen for the pySDC library, that may follow Python standards (or not ...), detailed in the [naming conventions](./docs/contrib/03_naming_conventions.md) page.
 
 Finally, while `pySDC` provides many base functionalities that implement classical flavors of SDC, it also allows problem-specific applications through Object-Oriented Programming (OOP) and the implementation of custom inherited classes.
-This follows a specific OOP framework, you can look at the page on [custom implementations](./docs/contrib/04_custom_implementations.md) for more details.
+This follows a specific OOP framework, you can look at the page on [custom implementations](./docs/contrib/04_custom_implementations.md) for more details. Additional guideline are also given on how to [document the code](./docs/contrib/05_documenting_code.md) in `pySDC`.
 
 1. [GitHub Forks and Pull Requests](https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/getting-started/about-collaborative-development-models)
 2. [Pull Requests Recommendations](./docs/contrib/01_pull_requests.md)
 3. [Continuous Integration](./docs/contrib/02_continuous_integration.md)
 4. [Naming Conventions](./docs/contrib/03_naming_conventions.md)
 5. [Custom Implementations](./docs/contrib/04_custom_implementations.md)
+6. [Documenting Code](./docs/contrib/05_documenting_code.md)
 
 :arrow_left: [Back to main page](./README.md)

docs/contrib/02_continuous_integration.md

@@ -56,7 +56,7 @@ This stage allows to checks how much of the `pySDC` code is tested by the previo
 - `pySDC/projects`
 - `pySDC/tutorial`
 
-This analysis is done in parallel to the test each time a pull is done on any branch (main repository or fork). 
+This analysis is done in parallel to the test each time a pull is done on any branch (main repository or fork).
 You can look at the current coverage report for the master branch [here](https://parallel-in-time.org/pySDC/coverage/index.html) or compare the results with previous builds [here](https://app.codecov.io/gh/Parallel-in-Time/pySDC). Codecov will also comment on any pull request, indicating the change of coverage.
 
 During developments, you can also run the coverage tests locally, using :
@@ -82,20 +82,18 @@ This will generate the coverage report in a `htmlcov` folder, and you can open t
 ### Coverage exceptions
 
 Some types of code lines will be ignored by the coverage analysis (_e.g_ lines starting with `raise`, ...), see the `[tool.coverage.report]` section in `pyproject.toml`.
-Part of code (functions, conditionaly, for loops, etc ...) can be ignored by coverage analysis using the `# pragma: no cover`, for instance 
+Part of code (functions, conditionaly, for loops, etc ...) can be ignored by coverage analysis using the `# pragma: no cover`, for instance
 
 ```python
 # ...
 # code analyzed by coverage
 # ...
-# pragma: no cover
-if condition:
+if condition:  # pragma: no cover
     # code ignored by coverage
 # ...
 # code analyzed by coverage
 # ...
-# pragma: no cover
-def function():
+def function():  # pragma: no cover
     # all function code is ignored by coverage
 ```
 
@@ -109,7 +107,7 @@ If you think the pragma should be used in other parts of your pull request, plea
 ## Documentation generation
 
 Documentation is built using [sphinx](https://www.sphinx-doc.org/en/master/).
-To check its generation, you can wait for all the CI tasks to download the `docs` artifacts, unzip it and open the `index.html` file there with you favorite browser. 
+To check its generation, you can wait for all the CI tasks to download the `docs` artifacts, unzip it and open the `index.html` file there with you favorite browser.
 
 However, when you are working on documentation (of the project, of the code, etc ...), you can already build and check the website locally :
 
@@ -124,7 +122,7 @@ sphinx-build -b html docs/source docs/build/html
 
 Then you can open `docs/build/html/index.html` using you favorite browser and check how your own documentation looks like on the website.
 
-> :bell: **Important** : running all the tests is necessary to generate graphs and images used by the website. 
+> :bell: **Important** : running all the tests is necessary to generate graphs and images used by the website.
 > But you can still generate the website without it: just all images for the tutorials, projects and playgrounds will be missing.
 > This approach can be considered for local testing of your contribution when it does not concern parts containing images (_i.e_ project or code documentation).
 

docs/contrib/04_custom_implementations.md

@@ -1,7 +1,62 @@
 # Custom implementation guidelines
 
-... in construction ...
+`pySDC` solves (non-)linear ODE of the form
+
+$$
+\frac{du}{dt} = f(u, t), \quad u(0)=u_0, \quad t \in [0, T].
+$$
+
+where $u$ a vector or scalar, representing one or more variables.
+
+The type of variable, the definition of the right-hand side $f(u,t)$ and initial solution value $u_0$ are defined in a given `Problem` class.
+
+... to be continued ...
+
+## Implementing a custom problem class
+
+Any problem class inherit from the same base class, that is (currently) the `ptype` class from the module `pySDC.code.Problem`.
+Each custom problem should inherit from this base class, like for the following example template :
+
+- right-hand side $f(u,t)=\lambda u +  ct$ with
+    - $\lambda$ one or more complex values (in vector form)
+    - $c$ one scalar value
+
+```python
+import numpy as np
+
+from pySDC.core.Problem import ptype
+from pySDC.core.Errors import ProblemError
+from pySDC.implementations.datatype_classes.mesh import mesh
+
+class MyCustomProblem(ptype):
+    # 1) Provide datatype class as class attribute
+    dtype_u = mesh  # -> used for u values
+    dtype_f = mesh  # -> used for f(u,t) values
+
+    # 2) Define constructor
+    def __init__(self, lam, c):
+
+        # Store lambda values into a numpy array (with copy) + check
+        lam = np.array(lam)
+        if len(lam.shape) > 1:
+            raise ProblemError(f"lambda values must be given as 1D vector, got shape {lam.shape}")
+
+        # Call base constructor
+        super().__init__(init=(lam.size, None, lam.dtype))
+
+        # Register parameters
+        self._makeAttributeAndRegister('lam', 'c', localVars=locals(), readOnly=True)
+
+    # 3) Define RHS function
+    def eval_f(self, u, t):
+        f = self.f_init  # Generate new datatype to store result
+        f[:] = self.lam*u + self.c*t  # Compute RHS value
+        return f
+```
+
+:bell: The `_makeAttributeAndRegister` automatically add `lam` and `c`, and register them in a list of parameters that are printed in the outputs of `pySDC`.
+If you set `readOnly=True`, then those parameters cannot be changed after initializing the problem (if not specifies, use `readOnly=False`).
 
 :arrow_left: [Back to Naming Conventions](./03_naming_conventions.md) ---
 :arrow_up: [Contributing Summary](./../../CONTRIBUTING.md) ---
-:arrow_right: [Next to a cute picture of cat](https://www.vecteezy.com/photo/2098203-silver-tabby-cat-sitting-on-green-background)
+:arrow_right: [Next to Documenting Code](./05_documenting_code.md)

docs/contrib/05_documenting_code.md

@@ -0,0 +1,93 @@
+# Documenting Code
+
+When developing a new class or function, or improving current classes in `pySDC`, adding Python docstring to document the code is important, in particular :
+
+- to help developer understanding how classes and functions work when reading the code
+- to help user understanding how classes and functions work when reading the [documentation](https://parallel-in-time.org/pySDC/#api-documentation)
+
+`pySDC` follows the [NumPy Style Python Docstring](https://numpydoc.readthedocs.io/en/latest/format.html), below is simplified example
+for a class and a function :
+
+> :bell: Don't document the `init` function, but rather the class itself. Also describe parameters (given to the `__init__`) and attributes (stored into the class) separately.
+
+```python
+class LagrangeApproximation(object):
+    r"""
+    Class approximating any function on a given set of points using barycentric
+    Lagrange interpolation.
+
+    Let note :math:`(t_j)_{0\leq j<n}` the set of points, then any scalar
+    function :math:`f` can be approximated by the barycentric formula :
+
+    .. math::
+        p(x) =
+        \frac{\displaystyle \sum_{j=0}^{n-1}\frac{w_j}{x-x_j}f_j}
+        {\displaystyle \sum_{j=0}^{n-1}\frac{w_j}{x-x_j}},
+
+    where :math:`f_j=f(t_j)` and
+
+    .. math::
+        w_j = \frac{1}{\prod_{k\neq j}(x_j-x_k)}
+
+    are the barycentric weights.
+    The theory and implementation is inspired from `this paper <http://dx.doi.org/10.1137/S0036144502417715>`_.
+
+    Parameters
+    ----------
+    points : list, tuple or np.1darray
+        The given interpolation points, no specific scaling, but must be
+        ordered in increasing order.
+
+    Attributes
+    ----------
+    points : np.1darray
+        The interpolating points
+    weights : np.1darray
+        The associated barycentric weights
+    """
+
+    def __init__(self, points):
+        pass  # Implementation ...
+
+    @property
+    def n(self):
+        """Number of points"""
+        pass  # Implementation ...
+
+    def getInterpolationMatrix(self, times):
+        r"""
+        Compute the interpolation matrix for a given set of discrete "time"
+        points.
+
+        For instance, if we note :math:`\vec{f}` the vector containing the
+        :math:`f_j=f(t_j)` values, and :math:`(\tau_m)_{0\leq m<M}`
+        the "time" points where to interpolate.
+        Then :math:`I[\vec{f}]`, the vector containing the interpolated
+        :math:`f(\tau_m)` values, can be obtained using :
+
+        .. math::
+            I[\vec{f}] = P_{Inter} \vec{f},
+
+        where :math:`P_{Inter}` is the interpolation matrix returned by this
+        method.
+
+        Parameters
+        ----------
+        times : list-like or np.1darray
+            The discrete "time" points where to interpolate the function.
+
+        Returns
+        -------
+        PInter : np.2darray(M, n)
+            The interpolation matrix, with :math:`M` rows (size of the **times**
+            parameter) and :math:`n` columns.
+        """
+        pass  # Implementation ...
+
+```
+
+A more detailed example is given [here ...](https://sphinxcontrib-napoleon.readthedocs.io/en/latest/example_numpy.html)
+
+:arrow_left: [Back to custom implementations](./04_custom_implementations.md) ---
+:arrow_up: [Contributing Summary](./../../CONTRIBUTING.md) ---
+:arrow_right: [Next to a cute picture of cat](https://www.vecteezy.com/photo/2098203-silver-tabby-cat-sitting-on-green-background)

pySDC/core/Collocation.py

@@ -1,7 +1,5 @@
 import logging
-
 import numpy as np
-import scipy.interpolate as intpl
 
 from pySDC.core.Nodes import NodesGenerator
 from pySDC.core.Errors import CollocationError

pySDC/core/Common.py

@@ -0,0 +1,75 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Description
+-----------
+
+Module containing utility classe(s) from which inherit some of the pySDC base
+classes.
+"""
+from pySDC.core.Errors import ReadOnlyError
+
+
+class _MetaRegisterParams(type):
+    """Metaclass for RegisterParams base class"""
+
+    def __new__(cls, name, bases, dct):
+        obj = super().__new__(cls, name, bases, dct)
+        obj._parNamesReadOnly = set()
+        obj._parNames = set()
+        return obj
+
+
+class RegisterParams(metaclass=_MetaRegisterParams):
+    """
+    Base class to register parameters.
+
+    Attributes
+    ----------
+    params : dict (property)
+        Dictionary containing names and values of registered parameters.
+    _parNames : set of str
+        Names of all the registered parameters.
+    _parNamesReadOnly : set of str
+        Names of all the parameters registered as read-only.
+    """
+
+    def _makeAttributeAndRegister(self, *names, localVars=None, readOnly=False):
+        """
+        Register a list of attribute name as parameters of the class.
+
+        Parameters
+        ----------
+        *names : list of str
+            The name of the parameters to be registered (should be class attributes).
+        localVars : dict
+            Dictionary containing key=names and values=paramValues for each
+            parNames given in names. Can be provided, for instance, using
+            `locals()` built-in dictionary. MUST BE provided as soon as
+            names contains anything.
+        readOnly : bool, optional
+            Wether or not store the parameters as read-only attributes
+        """
+        if len(names) > 1 and localVars is None:
+            raise ValueError("a dictionary must be provided in localVars with parameters values")
+        # Set parameters as attributes
+        for name in names:
+            try:
+                super().__setattr__(name, localVars[name])
+            except KeyError:  # pragma: no cover
+                raise ValueError(f'value for {name} not given in localVars')
+        # Register as class parameter
+        if readOnly:
+            self._parNamesReadOnly = self._parNamesReadOnly.union(names)
+        else:
+            self._parNames = self._parNames.union(names)
+
+    @property
+    def params(self):
+        """Dictionary containing names and values of registered parameters"""
+        return {name: getattr(self, name) for name in self._parNamesReadOnly.union(self._parNames)}
+
+    def __setattr__(self, name, value):
+        if name in self._parNamesReadOnly:
+            raise ReadOnlyError(name)
+        super().__setattr__(name, value)

pySDC/core/Controller.py

@@ -185,12 +185,11 @@ def dump_setup(self, step, controller_params, description):
                     else:
                         out += '            %s = %s\n' % (k, v)
             out += '-->         Problem: %s\n' % L.prob.__class__
-            for k, v in vars(L.prob.params).items():
-                if not k.startswith('_'):
-                    if k in description['problem_params']:
-                        out += '-->             %s = %s\n' % (k, v)
-                    else:
-                        out += '                %s = %s\n' % (k, v)
+            for k, v in L.prob.params.items():
+                if k in description['problem_params']:
+                    out += '-->             %s = %s\n' % (k, v)
+                else:
+                    out += '                %s = %s\n' % (k, v)
             out += '-->             Data type u: %s\n' % L.prob.dtype_u
             out += '-->             Data type f: %s\n' % L.prob.dtype_f
             out += '-->             Sweeper: %s\n' % L.sweep.__class__

pySDC/core/Errors.py

@@ -68,3 +68,12 @@ class ProblemError(Exception):
     """
 
     pass
+
+
+class ReadOnlyError(Exception):  # pragma: no cover
+    """
+    Exception thrown when setting a read-only class attribute
+    """
+
+    def __init__(self, name):
+        super().__init__(f'cannot set read-only attribute {name}')

pySDC/core/Lagrange.py

@@ -74,6 +74,12 @@ class LagrangeApproximation(object):
     are the barycentric weights.
     The theory and implementation is inspired from `this paper <http://dx.doi.org/10.1137/S0036144502417715>`_.
 
+    Parameters
+    ----------
+    points : list, tuple or np.1darray
+        The given interpolation points, no specific scaling, but must be
+        ordered in increasing order.
+
     Attributes
     ----------
     points : np.1darray
@@ -83,14 +89,6 @@ class LagrangeApproximation(object):
     """
 
     def __init__(self, points):
-        """
-
-        Parameters
-        ----------
-        points : list, tuple or np.1darray
-            The given interpolation points, no specific scaling, but must be
-            ordered in increasing order.
-        """
         points = np.asarray(points).ravel()
 
         diffs = points[:, None] - points[None, :]

pySDC/core/Level.py

@@ -70,7 +70,7 @@ def __init__(self, problem_class, problem_params, sweeper_class, sweeper_params,
 
         # instantiate sweeper, problem and hooks
         self.__sweep = sweeper_class(sweeper_params)
-        self.__prob = problem_class(problem_params)
+        self.__prob = problem_class(**problem_params)
 
         # set level parameters and status
         self.params = _Pars(level_params)